About

"The program should, in some way, expand the concept of what computer paint programs are, as well as what mark making can be."

Recent posts

Omar Rizwan @omar

i feel like the C pointer arithmetic stuff is kind of a half-hearted abstraction, like it's not strong enough that you can just think 100% in terms of type sizes and avoid thinking about bytes
yeah i think memory allocation, disk I/O, and network I/O all kind of force this to break
because you have to give byte offsets and lengths to deal with some external system

i feel like the C pointer arithmetic stuff is kind of a half-hearted abstraction, like it's not strong enough that you can just think 100% in terms of type sizes and avoid thinking about bytes
yeah i think memory allocation, disk I/O, and network I/O all kind of force this to break
because you have to give byte offsets and lengths to deal with some external system

Omar Rizwan @omar

Hmm

calendar with Time: 07:60pm and Hour slider underneath that's slid all the way to the right. chat message "we're meeting at 7:60pm. put it in ur calendars"

calendar with Time: 07:60pm and Hour slider underneath that's slid all the way to the right. chat message "we're meeting at 7:60pm. put it in ur calendars"

Omar Rizwan @omar

https://www.johndcook.com/blog/2011/04/11/sliderules/

Omar Rizwan @omar

Cute

Debugging

Introduction

This document describes "Feathers", a debugger for compiled CHICKEN programs.

Debugging

Introduction

This document describes "Feathers", a debugger for compiled CHICKEN programs.

Omar Rizwan @omar

C and the push to have fixed numbers of slots

Osnr Today at 11:31 AM

i'm working on using generational indices for matches (which
currently have a finite limit on the C branch) and then fixing the
3 things from yesterday

an odd thing about the C branch is it now has fixed limits for
lots of things which i kind of pulled out of a hat (you can only
have 8 destructors per match, etc)

many of these limits are ok on current workload, but limit on (#
of match ids) and limit on (# of statement ids) will definitely
cause the system to have a max uptime, you'll just run out of
ids, so we need to use generational indices for thsoe

Osnr Today at 11:31 AM

i'm working on using generational indices for matches (which
currently have a finite limit on the C branch) and then fixing the
3 things from yesterday

an odd thing about the C branch is it now has fixed limits for
lots of things which i kind of pulled out of a hat (you can only
have 8 destructors per match, etc)

many of these limits are ok on current workload, but limit on (#
of match ids) and limit on (# of statement ids) will definitely
cause the system to have a max uptime, you'll just run out of
ids, so we need to use generational indices for thsoe

Omar Rizwan @omar

Olivia Kan-Sperling's piece in Spike Art Magazine about the physical computing work we've been doing: https://spikeartmagazine.com/?q=articles/discourse-emily-dickinson-apple-computer-house

Omar Rizwan @omar

i hate the trash can icon for deleting things... i don't want to think
about gross garbage in my pristine digital realm

i hate the trash can icon for deleting things... i don't want to think
about gross garbage in my pristine digital realm

Omar Rizwan @omar

"when apps get complex enough, the differences between CPUs and schedulers wash out in the chaos"

Omar Rizwan @omar

"and, trust me, the fans and hard drive do sound like they've been running for 15 years"

Omar Rizwan @omar

hmm https://news.ycombinator.com/item?id=25588664

screenshot of linked comment (against memory allocation/copying in receiving messages)

screenshot of linked comment (against memory allocation/copying in receiving messages)

Back to top Show older